Powerset - Is It Man or Mouse?

I received my much anticipated invitation to Powerset's Powerlabs the other day, and was excited to finally get an inside peek one of the years most interesting startups. I have covered Powerset many times since my first article back in January. The brainchild of Barney Pell, Steve Newcomb and Lorenzo Thione held great potential in providing natural language search relevance for the Web. Powerlabs does indicate some significant developments toward natural search, but the larger issue to me is: "Where is the company heading?" Barney Pell stepping down and Newcomb leaving indicates a little more than a "shift" in the company's direction and ultimately its success.

Have To and Want To

I have spoken with Barney on a couple of occasions, and just so the reader knows - this is a brilliant, honorable and warm individual. The reason I mention this is because I believe that Barney and Steve, like many scientists, got caught up in what we might call a "hard money" scenario in funding this potentially Web changing endeavor. Without going into specifics, great and altruistic scientists have been put in "crunch" situations since the invention of fire, and I know several others experiencing at least some pressure even now. In the final analysis a company's essence if you will, may be determined more by who's money is behind it than the great vision of the innovator.

Powerlabs

Us "Powerlabbers" were supposed to be cranking on Powerlabs back in September, but given the detail and complexity of what Powerset has been trying to do - combined with the obvious work that went into Powerlabs itself - this is somewhat understandable. Powerlabs essentially focuses the power of natural language search onto Wikipedia. I found this to be a good way to incrementally apply and test Powerset as well as being a good marketing tool. Powerlabs is fun, cute, high tech and ultimately pits semantic search against conventional types in a very effective way. Powermouse is a very well done graphical personalization and actual searches via Arts are revealing. Is a refined, cute and effective preemptive interface indicative of success in natural language search though?

Powerset's progress evident in comparative search

Powermouse illustrates indexing

People and Machines

I have always been an advocate of people being the driving force behind innovation and success. Certainly, enough money or other assets shoved into the vacuum can offset the people equation as a matter or practical physics and etc. - but in many cases the money dynamic often negatively affects the great people behind creativity. If Barney, Steve and Lorenzo had all needed funding without outside intervention, I would have given them a 90 percent chance of making Powerset everything it could be even 3 months ago. This is not to say the company won't achieve a level of success, but at least for Steve - this seems a matter of little consequence now. Obviously, Barney stepping down is significant as well, though he gives an honest and logical explanation here. If my sense about this is right, I really hate to see these brilliant and productive people put into a situation where their art may be compromised because of money.

Ultimately the Comparison

The comparison between Powerset and hakia has always existed, though I honestly never really intended to set these two innovations against one another. The inevitable first took shape in an article I did for ReadWriteWeb back in July, where I figuratively compared the two after talking with Barney about his work. In that comparison I also symbolically made Powerset into one sports car as opposed to hakia as another. As it turns out, the irony appears to be strikingly true in that hakia seems to be backed by a much more patient and "old world" group (like Ferrari's heritage), while Powerset's group might be the hard driving, Motor City, high profile, glitter type (like the Ford Shelby GT 500).

This is a rather normative approach to explaining these happenings but the screens I provide do tell a story that coincides with the apparent reality. At this stage, given the startups we have seen - and also the backing behind them (Dmitry Shapiro and Veoh as a great example of "soft money"), it is evident that Powerset has some problems at least. A release in 2009, unless it is AI of the most acutely intuitive kind, will certainly put the development behind the 8-ball even further. Hakia has already made substantial progress and in my view will be too far ahead to catch in the medium term. If this is the case, and Powerset's investors are as seemingly impatient as it appears to me, it is no wonder there is a shakeup going on. I hate to even point to this type of conclusion because I really want Barney to succeed with something that would help us all.

Hakia's rather undersated and simple hierarchy - demonstrative of patient confidence

3 of the 9 profiles of "me" generation investors - perhaps "me" says a lot?